See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/278728806
SOME PROPERTIES OF COMBINATION OF SOLUTIONS TO SECOND-ORDER
LINEAR DIFFERENTIAL EQUATIONS WITH ANALYTIC COEFFICIENTS OF
[P;Q]-ORDER IN THE UNIT DISC
Article in International Journal of Applied Mathematics and Computer Science · January 2014 CITATION
1
READS 79
2 authors:
Some of the authors of this publication are also working on these related projects:
On the Oscillation of Meromorphic Functions and Their Difference PolynomialsView project
Growth of solutions of linear differential equations in the complex planeView project Benharrat Belaidi
Université Abdelhamid Ibn Badis Mostaganem
127 PUBLICATIONS 244 CITATIONS SEE PROFILE
Latreuch Zinelaabidine
Université Abdelhamid Ibn Badis Mostaganem
45 PUBLICATIONS 79 CITATIONS SEE PROFILE
All content following this page was uploaded by Latreuch Zinelaabidine on 20 June 2015. The user has requested enhancement of the downloaded file.
ISSN: 1857-9574 , UDC: 517.537:517.929.2
SOME PROPERTIES OF COMBINATION OF SOLUTIONS TO SECOND-ORDER LINEAR DIFFERENTIAL EQUATIONS WITH ANALYTIC COEFFICIENTS OF
[P; Q]-ORDER IN THE UNIT DISC
BENHARRAT BELAÏDI1 AND ZINELÂABIDINE LATREUCH
Abstract. In this paper, we consider some properties on the growth and oscillation of combination of solutions of the linear dierential equation
f00+ A (z) f0+ B (z) f = 0;
with analytic coecients A (z) and B (z) with [p; q] order in the unit disc = fz 2 C : jzj < 1g.
1. Introduction and preliminaries In the year 2000, Heittokangas rstly investigated the growth and oscillation theory of complex dierential equation
(1.1) f(k)+ A
k 1(z) f(k 1)+ + A0(z) f = 0;
where A0(z) ; ; Ak 1(z) are analytic functions in
the unit disc (see, [15]). It is well-known that all solu-tions of (1.1) are analytic funcsolu-tions (see, [15]). After him many authors (see, [4], [5], [8], [9], [10], [11], [12], [13], [16], [22]) have investigated the complex dier-ential equation (1.1) and the second-order dierdier-ential equations
(1.2) f00+ A (z) f0+ B(z)f = 0;
(1.3) f00+ A (z) f = 0
with analytic and meromorphic coecients in the unit disc : In ([17], [18]), Juneja and his co-authors investigated some properties of entire functions of [p; q] order, and obtained some results concerning their growth. Later, Liu, Tu and Shi; Xu, Tu and Xuan; Li and Cao; Belaïdi; Latreuch and Belaïdi applied the concepts of entire (meromorphic) func-tions in the complex plane and analytic funcfunc-tions in the unit disc = fz 2 C : jzj < 1g of [p; q] order to investigate the complex dierential equation (1.1) (see [6], [7], [22], [23], [24], [26]). In this paper, we will
use this concept to study the growth and the oscilla-tion of the combinaoscilla-tion of two linearly independent
solutions f1and f2 of equation (1.2) in the unit disc.
In this paper, we assume that the reader is famil-iar with the fundamental results and the standard notations of the Nevanlinna's theory on the complex plane and in the unit disc = fz 2 C : jzj < 1g ; see ([14], [15], [19], [20], [25]).
In the following, we will give similar denitions as in ([17], [18]) for analytic and meromorphic func-tions of [p; q]-order, [p; q]-type and [p; q]-exponent of convergence of the zero-sequence in the unit disc. Denition 1.1. ([6],[22] ) Let p q 1 be inte-gers, and let f be a meromorphic function in ; the [p; q]-order of f (z) is dened by
[p;q](f) = lim sup r!1 log+p T (r; f) logq 1 1 r ;
where T (r; f) is the Nevanlinna characteristic function of f: For an analytic function f in , we also dene M;[p;q](f) = lim sup r!1 log+p+1M (r; f) logq 1 1 r ; where M (r; f) = max jzj=rjf (z)j :
Remark 1.1. It is easy to see that 0 [p;q](f)
+1 (0 M;[p;q](f) +1 ); for any p
q 1: By Denition 1.1, we have that [1;1] =
1corresponding author
2010 Mathematics Subject Classication. 34M10, 30D35.
Key words and phrases. Analytic functions, linear dierential equations, [p; q] order, [p; q] exponent of convergence of the sequence of distinct zeros, unit disc.
12 B. BELAÏDI AND Z. LATREUCH
(f) (M;[1;1] = M(f)) and [2;1] = 2(f)
M;[2;1]= M;2(f).
For the relationship between [p;q](f) and
M;[p;q](f) we have the following double inequality.
Proposition 1.1. ([6]) Let p q 1 be integers, and let f be an analytic function in of [p; q]-order.
(i) If p = q 1; then
[p;q](f) M;[p;q](f) [p;q](f) + 1:
(ii) If p > q 1, then
[p;q](f) = M;[p;q](f) :
Denition 1.2. ([22] ) Let p q 1 be integers. The [p; q]-type of a meromorphic function f (z) in of [p; q]-order (0 < < +1 ) is dened by [p;q](f) = lim sup r!1 log+ p 1T (r; f) logq 1 1 1 r :
Denition 1.3. ([22]) Let p q 1 be inte-gers. The [p; q]-exponent of convergence of the zero-sequence of f (z) in is dened by [p;q](f) = lim sup r!1 log+p Nr;1 f logq 1 1 r ; where Nr;1 f
is the integrated counting function of zeros of f (z) in fz : jzj rg. Similarly, the [p; q]-exponent of convergence of the sequence of distinct zeros of f (z) in is dened by
[p;q](f) = lim sup r!1 log+p Nr;1 f logq 1 1 r ; where Nr;1 f
is the integrated counting function of distinct zeros of f (z) in fz : jzj rg.
The study of the properties of linearly indepen-dent solutions of complex dierential equations is an old problem. In ([2], [3]), Bank and Laine obtained
some results about the product E = f1f2 of two
lin-early independent solutions f1 and f2 of (1.3) in the
complex plane. In [21], the authors have investigated the relations between the polynomial of solutions of (1.2) and small functions in the complex plane. They
showed that w = d1f1+ d2f2keeps the same
proper-ties of growth and oscillation of fj (j = 1; 2) ; where
f1 and f2 are two linearly independent solutions of
(1.2) and obtained the following results.
Theorem 1.1. ([21]) Let A (z) and B (z) be entire functions of nite order such that (A) < (B) and (A) < (B) < +1 if (B) = (A) >
0. Let dj(z) (j = 1; 2) be entire functions
that are not all vanishing identically such that
maxf (d1) ; (d2)g < (B). If f1 and f2 are two
nontrivial linearly independent solutions of (1.2),
then the polynomial of solutions w = d1f1+ d2f2
satises
(w) = (f1) = (f2) = +1
and
2(w) = (B) :
In the same paper, the authors studied also the zeros of the dierence between the polynomial of
so-lutions w = d1f1+ d2f2and entire functions of nite
order.
The remainder of the paper is organized as fol-lows. In Section 2, we shall show our main results which improve and extend many results in the above-mentioned papers. Section 3 is for some lemmas and basic theorems. The other sections are for the proofs of our main results.
2. Main results
A natural question arises: What can be said about similar situations in the unit disc for equation (1.2) in the terms of [p; q] order? Before we state our re-sults, we dene h and (z) by
h = H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 ; where H1= d1; H2= 0; H3= d2; H4= 0; H5= d01; H6= d1; H7= d02; H8= d2; H9= d001 d1B; H10= 2d01 d1A; H11= d002 d2B; H12= 2d02 d2A; H13= d1(3) 3d01B +d1AB d1B0; H14= 3d001 2d01A d1B + d1A2 d1A0; H15= d(3)2 3d02B + d2AB d2B0; H16= 3d002 2d02A d2B + d2A2 d2A0 and (z) = 2 d1d2d02 d22d01 h '(3)+ 2'00+ 1'0+ 0';
where ' 6 0, dj(j = 1; 2) are analytic functions of
nite [p; q]-order in and
(2.1) 2= 2 d1d2d 0 2 d22d01 A 3d1d2d002+ 3d22d001 h ; 1= 6d2(d 0 1d002 d02d001) + 2d2(d1d02 d2d01) B h (2.2) +2d2(d1d02 d2d01) A0+ 3d2(d2d001 d1d002) A h ;
0 = h1[(d1d02d002 3d2d02d001+ 2d2d01d002) A + 4d1(d02)2+ 3d22d001 3d1d2d002 4d2d01d02 B +2 d2d01d02 d1(d02)2 A0 +2 d1d2d02 d22d01 B0+ 6(d0 2)2d001 2d1d02d0002 + 2d2d01d0002 (2.3) 3d2d001d002 6d01d02d002+ 3d1(d002)2]:
Theorem 2.1. Let p q 1 be integers, and let A (z) and B (z) be analytic functions in of
nite [p; q]-order such that [p;q](A) < [p;q](B)
and 0 < [p;q](A) < [p;q](B) < +1 if [p;q](B) =
[p;q](A) > 0. Let dj(z) (j = 1; 2) be analytic
func-tions that are not all vanishing identically such
that maxf[p;q](d1) ; [p;q](d2)g < [p;q](B). If f1
and f2 are two nontrivial linearly independent
so-lutions of (1.2), then the polynomial of soso-lutions
(2.4) w = d1f1+ d2f2
satises
[p;q](w) = [p;q](f1) = [p;q](f2) = +1
and
[p;q](B) [p+1;q](w) M;
where and in the following M =
maxM;[p;q](A) ; M;[p;q](B) : Furthermore, if
p > q 1; then
[p+1;q](w) = [p;q](B) :
Example 2.1. ([16]) For > 0; the
func-tions f1(z) = exp(exp (1 z) ) and f2(z) =
exp (1 z) exp(exp (1 z) ) are linearly
independent solutions of (1.2) satisfying
[1;1](f1) = [1;1](f2) = +1 and [2;1](f1) = [2;1](f2) = ; where A (z) = 2 exp (1 z) (1 z)+1 (1 z)+1 1 + 1 z and B (z) = 2exp 2(1 z) (1 z)2+2 :
It is clear that [1;1](A) = [1;1](B) and [1;1](A) <
[1;1](B) : Then, by Theorem 2.1 for any two
analytic functions di(z) (i = 1; 2) of nite order
[1;1](di) < +1 (i = 1; 2) that are not all vanishing
identically such that max[1;1](d1) ; [1;1](d2) <
[1;1](B), the combination w = d1f1+ d2f2 is of
innite order [1;1](w) = +1 and [2;1](w) = .
From Theorem 2.1, we can obtain the following result.
Corollary 2.1. Let p q 1 be integers, and let
fi(z) (i = 1; 2) be two nontrivial linearly
indepen-dent solutions of (1.2), where A (z) and B (z) 6 0 are analytic functions of nite [p; q] order in
such that [p;q](A) < [p;q](B) or [p;q](A) =
[p;q](B) > 0 and 0 < [p;q](A) < [p;q](B) < +1;
and let dj(z) (j = 1; 2; 3) be analytic functions in
satisfying
max[p;q](dj) : j = 1; 2; 3 < [p;q](B)
and
d2(z) f2+ d1(z) f1= d3(z) :
Then dj(z) 0 (j = 1; 2; 3) :
Theorem 2.2. Under the hypotheses of Theorem 2.1, let ' (z) 6 0 be an analytic function in
with nite [p; q] order such that (z) 6 0: If f1
and f2 are two nontrivial linearly independent
so-lutions of (1.2), then the polynomial of soso-lutions
w = d1f1+ d2f2 satises (2.5) [p;q](w ') = [p;q](w ') = [p;q](w) = +1 and [p;q](B) [p+1;q](w ') = (2.6) [p+1;q](w ') = [p+1;q](w) M: Furthermore, if p > q 1; then [p+1;q](w ') = [p+1;q](w ') (2.7) = [p+1;q](w) = [p;q](B) :
Theorem 2.3. Let p q 1 be integers, and let A (z) and B (z) be analytic functions in of
nite [p; q] order such that [p;q](A) < [p;q](B).
Let dj(z) ; bj(z) (j = 1; 2) be nite [p; q] order
analytic functions in such that d1(z) b2(z)
d2(z) b1(z) 6 0. If f1 and f2 are two nontrivial
linearly independent solutions of (1.2), then
[p;q] d1f1+ d2f2 b1f1+ b2f2 = +1 and [p;q](B) [p+1;q] d1f1+ d2f2 b1f1+ b2f2 M: Furthermore, if p > q 1; then [p+1;q] d1f1+ d2f2 b1f1+ b2f2 = [p;q](B) :
14 B. BELAÏDI AND Z. LATREUCH
3. Auxiliary lemmas
Lemma 3.1. ([14], [15], [25]) Let f be a meromor-phic function in the unit disc and let k 2 N: Then
m
r;ff(k)
= S (r; f) ;
where S (r; f) = Olog+T (r; f) + log 1
1 r
;
pos-sibly outside a set E1 [0; 1) withRE1 1 rdr < +1 :
Lemma 3.2. ([1], [15]) Let g : (0; 1) ! R and h : (0; 1) ! R be monotone increasing functions such that g (r) h (r) holds outside of an
excep-tional set E2 [0; 1) for which RE2 1 rdr < +1 .
Then there exists a constant d 2 (0; 1) such that if s (r) = 1 d (1 r) ; then g (r) h (s (r)) for all r 2 [0; 1):
Lemma 3.3. ([6]) Let p q 1 be integers.
If A0(z) ; ; Ak 1(z) are analytic functions of
[p; q] order in the unit disc ; then every solu-tion f 6 0 of (1.1) satises
[p+1;q](f) = M;[p+1;q](f)
maxM;[p;q](Aj) : j = 0; 1; ; k 1 :
Lemma 3.4. ([22]) Let p q 1 be integers. Let
Aj (j = 0; ; k 1) ; F 6 0 be analytic functions
in ; and let f (z) be a solution of the dierential equation f(k)+ A k 1(z) f(k 1)+ + A1(z) f0+ A0(z) f = F satisfying max[p;q](Aj) (j = 0; ; k 1) ; [p;q](F ) < [p;q](f) = +1: Then we have [p;q](f) = [p;q](f) = [p;q](f) and [p+1;q](f) = [p+1;q](f) = [p+1;q](f) :
Lemma 3.5. ([22]) Let p q 1 be integers, and let f and g be non-constant meromorphic func-tions of [p; q]-order in : Then we have
[p;q](f + g) max[p;q](f) ; [p;q](g)
and
[p;q](fg) max[p;q](f) ; [p;q](g) :
Furthermore, if [p;q](f) > [p;q](g) ; then we
ob-tain
[p;q](f + g) = [p;q](fg) = [p;q](f) :
Lemma 3.6. ([22]) Let p q 1 be integers, and let f and g be meromorphic functions of [p;
q]-order in such that 0 < [p;q](f) ; [p;q](g) <
+1 and 0 < [p;q](f) ; [p;q](g) < +1: Then, we have (i) If [p;q](f) > [p;q](g) ; then [p;q](f + g) = [p;q](fg) = [p;q](f) : (ii) If [p;q](f) = [p;q](g) and [p;q](f) 6= [p;q](g) ; then [p;q](f + g) = [p;q](fg) = [p;q](f) = [p;q](g) :
Lemma 3.7. ([22]) Let p q 1 be integers, and
let Aj(z) (j = 0; ; k 1) be analytic functions
in satisfying max[p;q](Aj) : j = 1; ; k 1 < [p;q](A0) : If f 6 0 is a solution of (1.1), then [p;q](f) = +1 and [p;q](A0) [p+1;q](f) maxM;[p;q](Aj) : j = 0; ; k 1 : Furthermore, if p > q 1; then [p+1;q](f) = [p;q](A0) :
Lemma 3.8. Let p q 1 be integers, and let A (z) and B (z) be analytic functions in of
-nite [p; q] order such [p;q](A) < [p;q](B). If f1
and f2 are two nontrivial linearly independent
so-lutions of (1.2), then f1 f2 is of innite [p; q]-order and [p;q](B) [p+1;q] f1 f2 M: Furthermore, if p > q 1; then [p+1;q] f1 f2 = [p;q](B) :
Proof. Suppose that f1 and f2 are two
nontriv-ial linearly independent solutions of (1.2). Since
[p;q](B) > [p;q](A), then by Lemma 3.7
[p;q](f1) = [p;q](f2) = +1; [p;q](B)
(3.1) [p+1;q](f1) = [p+1;q](f2) M:
Furthermore, if p > q 1; then
[p+1;q](f1) = [p+1;q](f2) = [p;q](B) :
On the other hand (3.2) f1 f2 0 = W (ff12; f2) 2 ;
where W (f1; f2) = f1f20 f2f10 is the Wronskian of
f1and f2: By using (1.2) we obtain that
W0(f
1; f2) = A (z) W (f1; f2);
which implies that
(3.3) W (f1; f2) = K exp(
Z
where RA(z)dz is the primitive of A (z) and K 2 Cn f0g : By (3.2) and (3.3) we have (3.4) f1 f2 0 = Kexp( R A(z)dz) f2 2 : Since [p;q](f2) = +1; [p+1;q](f2) [p;q](B) > [p;q](A) if p q 1 and [p+1;q](f2) = [p;q](B) >
[p;q](A) if p > q 1; then by using (3.1) and Lemma
3.5 we obtain from (3.4) [p;q] f1 f2 = [p;q](f2) = +1; [p;q](B) [p+1;q] f1 f2 = [p+1;q](f2) M if p q 1 and [p+1;q] f1 f2 = [p+1;q](f2) = [p;q](B) if p > q 1:
Lemma 3.9. ([6]) Let p q 1 be integers. Let f be a meromorphic function in the unit disc
such that [p;q](f) = < +1, and let k 1 be an
integer. Then for any " > 0; m r;ff(k) = O expp 1 ( + ") logq 1 1 r
holds for all r outside a set E3 [0; 1) with
R
E3
dr
1 r < +1 :
Lemma 3.10. Let p q 1 be integers, and let f be a meromorphic function in with [p; q]
-order 0 < [p;q](f) = < +1 and [p; q] - type
0 < [p;q](f) = < +1: Then for any given < ;
there exists a subset E4 of [0; 1) that has an
in-nite logarithmic measure RE4 dr
1 r = +1 such
that logp 1T (r; f) > hlogq 1 1
1 r
i
holds for
all r 2 E4:
Proof. By the denitions of [p; q] order and [p; q]
-type, there exists an increasing sequence frmg+1m=1
[0; 1) (rm! 1 ) satisfying m1 + 1 m1rm< rm+1 and lim m!+1 logp 1T (rm; f) logq 1 1 1 rm = :
Then there exists a positive integer m0such that for
all m m0and for any given 0 < " < ; we have
(3.5) logp 1T (rm; f) > ( ") logq 1 1 1 rm : For any given < "; there exists a positive integer
m1such that for all m m1we have
(3.6) 2 4logq 1 1 m1 1 1 r logq 1 1 1 r 3 5 > ":
Take m m2 = max fm0; m1g : By (3.5) and (3.6),
for any r 2 [rm;m1 + 1 m1rm]; we have
logp 1T (r; f) logp 1T (rm; f) > ( ") logq 1 1 1 rm ( ") logq 1 1 m1 1 1 r > logq 1 1 1 r : Set E4 = m=m+1[ 2 rm;m1 + 1 m1rm; then there holds mlE4= +1X m=m2 1 m+(1Zm1)rm rm dt 1 t = X+1 m=m2 logm 1m = +1: Lemma 3.11. Let p q 1 be integers, and let A (z) and B (z) be analytic functions in of
nite [p; q] order such [p;q](A) < [p;q](B) and
0 < [p;q](A) < [p;q](B) < +1 if [p;q](B) =
[p;q](A) > 0. If f 6 0 is a solution of (1.2), then
[p;q](f) = +1 ; [p;q](B) [p+1;q](f) M
and
[p+1;q](f) = [p;q](B)
if p > q 1.
Proof. If [p;q](B) > [p;q](A), then the result can be
deduced by Lemma 3.7. We prove only the case when
[p;q](B) = [p;q](A) = and [p;q](B) > [p;q](A) >
0: Since f 6 0; then by (1.2) B = f00 f + A f0 f :
Suppose that f is of nite [p; q] order [p;q](f) =
< +1: Then by Lemma 3.9 T (r; B) T (r; A)+O expp 1 ( + ") logq 1 1 r
holds for all r outside a set E3 [0; 1) withRE3 1 rdr <
+1; which implies by using Lemma 3.2 the contra-diction
[p;q](B) [p;q](A) :
Hence [p;q](f) = +1: By Lemma 3.3, we have
[p+1;q](f) = M;[p+1;q](f)
maxM;[p;q](A) ; M;[p;q](B) :
On the other hand, since [p;q](f) = +1; then by
Lemma 3.1 (3.7) T (r; B) T (r; A)+O log+T (r; f) + log 1 1 r
16 B. BELAÏDI AND Z. LATREUCH
holds for all r outside a set E1 [0; 1) with
R
E1
dr
1 r < +1. By [p;q](B) > [p;q](A) > 0, we
choose 0; 1 satisfying [p;q](B) > 0 > 1 >
[p;q](A) such that for r ! 1 ; we have
(3.8) T (r; A) expp 1 1 logq 1 1 1 r :
By Lemma 3.10, there exists a subset E4 [0; 1) of
innite logarithmic measure such that
(3.9) T (r; B) > expp 1 0 logq 1 1 1 r :
By (3.7)-(3.9) we obtain for all r 2 E4nE1
expp 1 0 logq 1 1 1 r expp 1 1 logq 1 1 1 r (3.10) +O log+T (r; f) + log 1 1 r : By using (3.10) and Lemma 3.2, we obtain
[p;q](B) [p+1;q](f) : Hence [p;q](B) [p+1;q](f) maxM;[p;q](A) ; M;[p;q](B) if p q 1 and [p+1;q](f) = [p;q](B) if p > q 1: 4. Proof of Theorem 2.1
Proof. In the case when d1(z) 0 or d2(z) 0; then
the conclusions of Theorem 2.1 are trivial. Suppose
that f1and f2are two nontrivial linearly independent
solutions of (1.2) and dj(z) 6 0 (j = 1; 2) : Then by
Lemma 3.11, we have
[p;q](f) = +1; [p;q](B) [p+1;q](f) M
if p q 1 and
[p+1;q](f) = [p;q](B)
if p > q 1. Suppose that d1 = cd2; where c is a
complex number: Then, we obtain
w = d1f1+ d2f2= cd2f1+ d2f2= (cf1+ f2) d2:
Since f = cf1 + f2 is a solution of (1.2) and
[p;q](d2) < [p;q](B) ; then we have
[p;q](w) = [p;q](cf1+ f2) = +1;
[p;q](B) [p+1;q](w) = [p+1;q](cf1+ f2) M
if p q 1 and
[p+1;q](w) = [p+1;q](cf1+ f2) = [p;q](B)
if p > q 1. Suppose now that d16 cd2 where c is a
complex number. Dierentiating both sides of (2.4), we obtain
(4.1) w0= d0
1f1+ d1f10 + d02f2+ d2f20:
Dierentiating both sides of (4.1), we obtain
(4.2) w00= d00 1f1+2d01f10+d1f100+d002f2+2d02f20+d2f200: Substituting f00 j = A (z) fj0 B (z) fj(j = 1; 2) into equation (4.2), we have w00= (d00 1 d1B) f1+ (2d01 d1A) f10 (4.3) + (d00 2 d2B) f2+ (2d02 d2A) f20:
Dierentiating both sides of (4.3) and by substituting f00 j = A (z) fj0 B (z) fj (j = 1; 2) ; we obtain w000=d(3) 1 3d01B + d1(AB B0) f1 + 3d00 1 2d01A + d1 A2 A0 Bf10 +d(3)2 3d0 2B + d2(AB B0) f2 (4.4) + 3d00 2 2d02A + d2 A2 A0 Bf20: By (2.4) and (4.1)-(4.4) we have 8 > > > > > > > > > > > > > < > > > > > > > > > > > > > : w = d1f1+ d2f2; w0= d0 1f1+ d1f10+ d02f2+ d2f20; w00= (d00 1 d1B) f1+ (2d01 d1A) f10 + (d00 2 d2B) f2+ (2d02 d2A) f20; w000=d(3) 1 3d01B + d1(AB B0) f1 + 3d00 1 2d01A + d1 A2 A0 Bf10 +d(3)2 3d0 2B + d2(AB B0) f2 + 3d00 2 2d02A + d2 A2 A0 Bf20:
To solve this system of equations, we need rst to prove that h 6 0: By simple calculations we obtain
h = H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 = 2 (d1d02 d2d01)2B + d2 2d01d001+ d12d02d002 d1d2d01d002 d1d2d02d001 A 2 (d1d02 d2d01)2A0+ 2d1d2d01d0002 + 2d1d2d02d0001 6d1d2d001d002 6d1d01d02d002 6d2d01d02d001 +6d1(d02)2d001+ 6d2(d10)2d002 2d22d01d0001 2d2 1d02d0002 + 3d21(d200)2+ 3d22(d001)2:
It is clear that (d1d02 d2d01)26 0 because d16= cd2.
Since
and (d1d02 d2d01)2 6 0; then by using Lemma 3.6
we can deduce that [p;q](h) = [p;q](B) > 0: Hence
h 6 0: By Cramer's method we have
f1= w H2 H3 H4 w0 H6 H7 H8 w00 H 10 H10 H12 w(3) H14 H15 H16 h (4.5) = 2 d1d2d02 d22d01 h w(3)+2w00+1w0+0w;
where j (j = 0; 1; 2) are meromorphic functions
in of nite [p; q]-order which are dened in
(2.1)-(2.3). By (4.5) and Lemma 3.5, we have
[p;q](f1) [p;q](w) ([p+1;q](f1) [p+1;q](w)) and
by (2.4) we have [p;q](w) [p;q](f1) ([p+1;q](w)
[p+1;q](f1)): Thus [p;q](w) = [p;q](f1) and
[p+1;q](w) = [p+1;q](f1) :
5. Proof of Corollary 2.1
Proof. We suppose there exists j = 1; 2; 3 such
that dj(z) 6 0 and we obtain a contradiction. If
d1(z) 6 0 or d2(z) 6 0; then by Theorem 2.1 we have
[p;q](d1f1+ d2f2) = +1 = [p;q](d3) < [p;q](B)
which is a contradiction. Now if d1(z) 0,
d2(z) 0 and d3(z) 6 0 we obtain also a
contra-diction. Hence dj(z) 0 (j = 1; 2; 3).
6. Proof of Theorem 2.2 Proof. By Theorem 2.1, we have
[p;q](w) = +1; [p;q](B) [p+1;q](w) M
if p q 1 and
[p+1;q](w) = [p;q](B)
if p > q 1. Set g (z) = d1f1 + d2f2 ':
Since [p;q](') < +1; then by Lemma 3.5 we
have [p;q](g) = [p;q](w) = +1; [p+1;q](g) = [p+1;q](w) : In order to prove [p;q](w ') = [p;q](w ') = +1 and [p+1;q](w ') = [p+1;q](w ') = [p+1;q](w) ; we need to prove only [p;q](g) = [p;q](g) = +1 and [p+1;q](g) = [p+1;q](g) = [p+1;q](w) : By w = g + ' we get from (4.5) (6.1) f1= 2 d1d2d 0 2 d22d01 h g(3)+ 2g00+ 1g0+ 0g + ; where = 2 d1d2d02 d22d01 h '(3)+ 2'00+ 1'0+ 0':
Substituting (6.1) into equation (1.2), we obtain 2 d1d2d02 d22d01 h g(5)+ 4 X j=0 jg(j) = ( 00+ A (z) 0+ B (z) ) = F (z) ;
where j(j = 0; ; 4) are meromorphic functions of
nite [p; q]-order in . Since 6 0 and [p;q]( ) <
+1; it follows that is not a solution of (1.2), which implies that F (z) 6 0: Then, by applying Lemma 3.4
we obtain (2.5), (2.6) and (2.7).
7. Proof of Theorem 2.3
Proof. Suppose that f1and f2are two nontrivial
lin-early independent solutions of (1.2). Then by Lemma 3.8, we have (7.1) [p;q] f1 f2 = +1; [p;q](B) [p+1;q] f1 f2 M if p q 1 and (7.2) [p+1;q] f 1 f2 = [p;q](B) if p > q 1. Set g = f1 f2: Then w (z) = db1(z) f1(z) + d2(z) f2(z) 1(z) f1(z) + b1(z) f2(z) =db1(z) g (z) + d2(z) 1(z) g (z) + b2(z):
By Lemma 3.5, it follows that
[p+1;q](w)
maxf[p+1;q](dj) ; [p+1;q](bj) ( j = 1; 2) ; [p+1;q](g)g
(7.3) = [p+1;q](g) :
On the other hand
g (z) = bb2(z) w (z) d2(z)
1(z) w (z) d1(z);
which implies by Lemma 3.5 that [p;q](w)
[p;q](g) = +1 and [p+1;q](g) maxf[p+1;q](dj) ; [p+1;q](bj) ( j = 1; 2) ; [p+1;q](w)g (7.4) = [p+1;q](w) : By using (7.1)-(7.4), we obtain [p;q](w) = [p;q](g) = +1; [p;q](B) [p+1;q](w) = [p+1;q](g) M if p q 1 and [p+1;q](w) = [p+1;q](g) = [p;q](B) if p > q 1.
18 B. BELAÏDI AND Z. LATREUCH
Acknowledgment
The authors are grateful to the referee for his/her valuable comments which lead to the improvement of this paper.
References
[1] S. Bank: General theorem concerning the growth of solutions of rst-order algebraic dierential equations, Compositio Math. 25 (1972), 61-70.
[2] S. Bank, I. Laine: On the oscillation theory of f00+
A (z) f = 0: where A is entire, Trans. Amer. Math. Soc. 273(1) (1982), 351363.
[3] S. Bank, I. Laine: On the zeros of meromorphic solu-tions of second-order linear dierential equasolu-tions, Com-ment. Math. Helv. 58(4) (1983), 656677.
[4] B. Belaïdi:Oscillation of fast growing solutions of lin-ear dierential equations in the unit disc, Acta Univ. Sapientiae Math. 2(1) (2010), 2538.
[5] B. Belaïdi: Growth of solutions of linear dierential equations in the unit disc, Bull. Math. Analy. Appl., 3(1) (2011), 1426.
[6] B. Belaïdi: Growth of solutions to linear dierential equations with analytic coecients of [p,q]-order in the unit disc, Electron. J. Dierential Equations 156 (2011), 111.
[7] B. Belaïdi: On the [p,q]-order of analytic solutions of linear dierential equations in the unit disc, Novi Sad J. Math. 42(1) (2012), 117129.
[8] T. B. Cao, H. X. Yi:The growth of solutions of linear dierential equations with coecients of iterated order in the unit disc, J. Math. Anal. Appl. 319(1) (2006), 278 294.
[9] T. B. Cao: The growth, oscillation and xed points of solutions of complex linear dierential equations in the unit disc, J. Math. Anal. Appl. 352(2) (2009), 739-748. [10] T. B. Cao, H. X. Yi: On the complex oscillation
the-ory of f00+ Af = 0 where A(z) is analytic in the unit
disc, Math. Nachr. 282(6) (2009), 820-831.
[11] T. B. Cao, C. X. Zhu, K. Liu: On the complex os-cillation of meromorphic solutions of second order lin-ear dierential equations in the unit disc, J. Math. Anal. Appl. 374(1) (2011), 272281.
[12] Z. X. Chen, K. H. Shon: The growth of solutions of dierential equations with coecients of small growth in the disc, J. Math. Anal. Appl. 297(1) (2004), 285304. [13] A. El Farissi, B. Belaïdi, Z. Latreuch: Growth
and oscillation of dierential polynomials in the unit disc, Electron. J. Di. Equ., 87 (2010), 1-7.
[14] W. K. Hayman: Meromorphic functions, Clarendon Press, Oxford, 1964.
[15] J. Heittokangas: On complex dierential equations in the unit disc, Ann. Acad. Sci. Fenn. Math. Diss. 122 (2000), 1-54.
[16] J. Heittokangas, R. Korhonen, J. Rättyä: Fast growing solutions of linear dierential equations in the unit disc, Results Math. 49(3-4) (2006), 265278. [17] O. P. Juneja, G. P. Kapoor, S. K. Bajpai: On the
(p; q)-order and lower (p; q)-order of an entire function, J. Reine Angew. Math. 282 (1976), 53-67.
[18] O. P. Juneja, G. P. Kapoor, S. K. Bajpai: On the (p; q)-type and lower (p; q)-type of an entire function, J. Reine Angew. Math. 290 (1977), 385-405.
[19] I. Laine: Nevanlinna theory and complex dierential equations, de Gruyter Studies in Mathematics, 15, Walter de Gruyter & Co., Berlin-New York, 1993.
[20] I. Laine: Complex dierential equations, Handbook of Dierential Equations: Ordinary Dierential Equations, 4 (2008), 269-363.
[21] Z. Latreuch, B. Belaïdi: Growth and oscillation of some polynomials generated by solutions of complex dif-ferential equations, Opuscula Math. 35(1) (2015), 8598. (In Press).
[22] Z. Latreuch, B. Belaïdi: Linear dierential equa-tions with analytic coecients of [p,q]-order in the unit disc, Sarajevo J. Math. 9(1) (2013), 7184.
[23] L. M. Li, T. B. Cao: Solutions for linear dierential equations with meromorphic coecients of [p; q]-order in the plane, Electron. J. Dierential Equations 195 (2012), 115.
[24] J. Liu, J. Tu, L. Z. Shi: Linear dierential equa-tions with entire coecients of [p, q]-order in the complex plane, J. Math. Anal. Appl. 372 (2010), 5567.
[25] M.Tsuji: Potential theory in modern function theory, Reprinting of the 1959 original. Chelsea Publishing Co., New York, 1975.
[26] H. Y. Xu, J. Tu, Z. X. Xuan: The oscillation on solu-tions of some classes of linear dierential equasolu-tions with meromorphic coecients of nite [p; q]-order, Sci. World J., Volume 2013 (2013), Article ID 243873, 8 pages. Department of Mathematics
Laboratory of Pure and Applied Mathematics University of Mostaganem (UMAB)
B. P. 227 Mostaganem, Algeria E-mail address: belaidi@univ-mosta.dz Department of Mathematics
Laboratory of Pure and Applied Mathematics University of Mostaganem (UMAB)
B. P. 227 Mostaganem, Algeria E-mail address: z.latreuch@gmail.com
View publication stats View publication stats